[ANSYS Meshing] 2D inflation layer contains triangles

jones007 · May 23, 2016, 21:20

I'm trying to build a mesh around a thin airfoil using an inflation layer and curvature control. Under curvature control I cannot seem to input curvature normal angle values of less that 0.1 degree. If I use this, then there seems to be a limit for the initial wall spacing height in the inflation layer. If I use a relatively large first height, then I get a well behaved inflation region - all quads visible in the 2D plane. If I make the first layer a little thinner, than most of the inflation layer turns to thin triangular cells - basically each quad cut in half diagonally. I thought perhaps this was a graphic defect, but the solution on this grid will not converge. Solutions on the grid with a thicker first layer converge nicely, but the Yplus value is too large.

With experimentation, I found that there was some form of coupling between the curvature normal angle and the first inflation layer height. I don't think it is a direct coupling, but the curvature normal angle sets up the circumferential spacing around the foil, and this seems to be coupled to the acceptable first cell height.

Does anyone know how to set a smaller curvature normal angle or perhaps another way to handle this? Thanks.

sheaker · March 31, 2019, 14:19

Dear All,

I am facing the same issue. Any solution?

Best regards,
Oskar

sy yao · April 6, 2020, 09:18

I faced the similar problem. I applied same settings for both global and local inflation. I tried varying some parameters such as number of divisions along the airfoil and total thickness of inflation, the problem persists. I'm using ANSYS 19.2.

I hope I can find the solution asap, this is for my fyp (it's kind of urgent).

Thanks in advance!

Gert-Jan · April 6, 2020, 09:54

It is best to extend the tail with a thin surface through your fluid domain and then inflate both sides. As a result the inflation won't degrade at the end of the tail but will continue away from the tail. Extend the thin surface to the outlet.
Then in the CFD setup, use your thin surface as an interface through which the fluid can penetrate.

sy yao · May 1, 2020, 10:27

Thank you for your reply, Gert-Jan.

Do you mean split the domain at the trailing edge of the airfoil like this?

Gert-Jan · May 1, 2020, 10:36

Quote:

Originally Posted by sy yao

Thank you for your reply, Gert-Jan.

Do you mean split the domain at the trailing edge of the airfoil like this?

Yes. This is what I meant.

sy yao · May 4, 2020, 02:14

Thanks, Gert-Jan.

I have solved the inflation meshing problem.

My mesh will roughly be something as shown in the picture attached.

The blue region is my rotating domain while the grey one is static. Will the big difference between the size of elements has negative impact on the cfd results later?

Gert-Jan · May 4, 2020, 03:16

Impossible to say. I don't know what question you are trying to answer using CFD.
Now I would say: give it a try and see what it brings.

sy yao · May 4, 2020, 08:42

Okay, I will give it a try.

Thanks again, Gert-Jan.

May 23, 2016, 21:20	2D inflation layer contains triangles	#1
jones007 New Member Kevin Jones Join Date: Feb 2011 Posts: 17 Rep Power: 15	I'm trying to build a mesh around a thin airfoil using an inflation layer and curvature control. Under curvature control I cannot seem to input curvature normal angle values of less that 0.1 degree. If I use this, then there seems to be a limit for the initial wall spacing height in the inflation layer. If I use a relatively large first height, then I get a well behaved inflation region - all quads visible in the 2D plane. If I make the first layer a little thinner, than most of the inflation layer turns to thin triangular cells - basically each quad cut in half diagonally. I thought perhaps this was a graphic defect, but the solution on this grid will not converge. Solutions on the grid with a thicker first layer converge nicely, but the Yplus value is too large. With experimentation, I found that there was some form of coupling between the curvature normal angle and the first inflation layer height. I don't think it is a direct coupling, but the curvature normal angle sets up the circumferential spacing around the foil, and this seems to be coupled to the acceptable first cell height. Does anyone know how to set a smaller curvature normal angle or perhaps another way to handle this? Thanks.

March 31, 2019, 14:19		#2
sheaker Senior Member Oskar Join Date: Nov 2015 Location: Poland Posts: 184 Rep Power: 10	Dear All, I am facing the same issue. Any solution? Best regards, Oskar __________________ How to set two stroke engine with scavenging? Salome, sliding interface and "Zero length edge detected."

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Total Thickness of Inflation Layer	prikeyma	CFX	5	August 28, 2020 03:22
[snappyHexMesh] snappyHexMesh Boundary Layer at Corner	panpanzhong	OpenFOAM Meshing & Mesh Conversion	5	July 3, 2018 05:53
Prismatic boundary layer	KateEisenhower	enGrid	5	September 15, 2015 07:48
[snappyHexMesh] Boundary layer generation problems	ivan_cozza	OpenFOAM Meshing & Mesh Conversion	0	October 6, 2010 13:47
errors	Fahad	Main CFD Forum	0	March 23, 2004 13:20

April 6, 2020, 09:54		#4
Gert-Jan Senior Member Gert-Jan Join Date: Oct 2012 Location: Europe Posts: 1,827 Rep Power: 27	It is best to extend the tail with a thin surface through your fluid domain and then inflate both sides. As a result the inflation won't degrade at the end of the tail but will continue away from the tail. Extend the thin surface to the outlet. Then in the CFD setup, use your thin surface as an interface through which the fluid can penetrate.

May 4, 2020, 03:16		#8
Gert-Jan Senior Member Gert-Jan Join Date: Oct 2012 Location: Europe Posts: 1,827 Rep Power: 27	Impossible to say. I don't know what question you are trying to answer using CFD. Now I would say: give it a try and see what it brings.

May 4, 2020, 08:42		#9
sy yao New Member Join Date: Mar 2020 Posts: 5 Rep Power: 6	Okay, I will give it a try. Thanks again, Gert-Jan.